Induction of an immune response to an antigen and the magnitude of that response depend upon a complex interplay among the antigen, various types of immune cells, and co-stimulatory molecules including cytokines. The timing and extent of exposure of the immune cells to the antigen and the co-stimulatory milieu further modulate the immune response. Within the body, these various cell types and additional factors are brought into proximity in lymphoid tissue such as lymph nodes. Of the numerous cell types involved in the process, antigen-presenting cells (APC), such as macrophages and dendritic cells, transport antigen from the periphery to local, organized lymphoid tissue, process the antigen and present antigenic peptides to T cells as well as secrete co-stimulatory molecules. Thus, if antigen reaches lymph organs in a localized staggered manner, presenting antigenic epitopes, under the optimal concentration gradient and under the appropriate environment comprising co-stimulatory molecules, a response is induced in the draining lymph node.
In this manner, a foreign antigen introduced into the body, such as by means of a vaccination, may or may not result in the development of a desirably robust immune response. Antigens used for vaccination include attenuated and inactivated bacteria and viruses and their components. The success of vaccination depends in part on the type and quantity of the antigen, the location of the site of immunization, and the status of the immune system at the time of vaccination. Not all antigens are equally immunogenic, and for poorly immunogenic antigens, there are few alternatives available to increase the effectiveness of the immunization. Whereas in experimental animals numerous techniques are available to enhance the development of the immune response, such as conjugating the antigen to a more immunogenic carrier protein or biomolecule (e.g., keyhole limpet hemocyanin), or the use of adjuvants such as Freund's Adjuvant or Ribi. For human vaccinations such techniques and adjuvants are not available. Thus, numerous diseases that would otherwise be preventable by vaccination before exposure to the infectious agent, or in the case of a therapeutic vaccine, that may induce the development of an effective immune response to an existing disease-causing agent or cell, such as cancer, are not available to the patient.
Sponge implant studies have been performed in mammals to assess the immune cell population attracted to a foreign body, which produce what is called a sterile abscess, and sponges prior to or after implantation have been loaded with antigen to further study the attracted cell population. Vallera et al. (1982, Cancer Research 42:397-404) implanted sponges containing tumor cells in mice to examine the composition of cells attracted over a 16 day period, and found that at an early time, cytotoxic cell precursors were present, and cytotoxicity peaked at day 16. Sponges containing tumor cells implanted in mice that had been previously immunized with tumor cells showed a more rapid appearance of cytotoxic cells in the sponge. In neither case did cells from the spleen, lymph nodes or peritoneum show cytotoxicity, which suggested a highly localized response to the antigen in the sponge. Zangemeister-Wittke et al. (1989, J. Immunol. 143:379-385) injected a tumor vaccine into sponges implanted in tumor-immune mice, and monitored the generation of a secondary immune response at the sponge site. No accompanying effect was apparent in lymph nodes adjacent to the implanted sponge.
Other devices which overcome some of the limitations of sponges for immunomodulation have been proposed. U.S. Pat. No. 4,919,929 teaches that an antigen can be loaded into solid shaped particles, which slowly release the antigen following implantation. This type of device is envisaged to increase the antibody titers in the milk of mammals and thereby confer higher levels of immunity in those who consume it. WO application 93/17662 describes a device that consists of an impervious membrane surrounding a core, which is a gel loaded with a therapeutically active ingredient (including antigens). There is at least one port in the impervious membrane that is capable of releasing the active to the surroundings. The use of the membrane is shown to slow the rate of release of the bioactive molecule (including antigens) relative to the gel alone. This device therefore primarily serves as a reservoir for slow release and does not facilitate the interaction of cells with the bioactive, which necessarily must occur outside of the device. In U.S. Pat. No. 4,732,155, a device is proposed where there is a reservoir that provides prolonged release of a chemoattractant, which is surrounded by a web of fibers adjacent to the reservoir. Cells are attracted to the reservoir and become trapped in the fibrous web. This device is proposed for use in characterizing allergic and inflammatory responses to test compounds by allowing controlled exposure to the compound and by trapping the cells that respond to it. This device both incorporates a mechanism for prolonged exposure to an antigen as well as a mechanism to facilitate cellular interaction with the antigen. The open web of fibers in this device; however, does not enable local retention of the cytokines and chemokines being secreted by the responding cells since an open web of fibers will not provide diffusional resistance to soluble factors.
This design is improved upon in WO 99/44583 which proposes a porous matrix which is housed in a perforated but otherwise impervious membrane. Antigen is loaded within the device and can be present either as native antigen or can be encapsulated in a slow releasing polymer that provides prolonged presentation of the antigen. Specific cells are attracted to the device by diffusion of the antigen from the perforations in the device and are also able to enter the device through the perforations, but the membrane provides sufficient diffusional resistance that cytokines secreted by cells become locally concentrated within the device. The high local densities of cells and cytokines produce a much more robust immune response than is seen with an uncontained matrix or with simple prolonged release to surrounding tissues.
The preferred embodiment of the device mentioned above envisages the porous matrix to be a sponge and the membrane to be a perforated tube. While very favorable immunomodulation is seen with the device, it is impractical to miniaturize and manufacture in large quantities. The primary reason is that it is very difficult to load a porous sponge into tubing. Sponges, due to their low bulk densities are mechanically weak and tend to tear easily when subjected to the tensile and compressive forces of loading into small diameter tubing. By reducing the bulk density, more favorable mechanical properties can be encountered however the matrix does not contain sufficient porosity to attain high cell densities. In addition, it is very difficult to cut small cylindrical cores of porous sponges for loading into tubes. The reason is that the poor mechanical properties of the porous sponge lead to tearing when the size of the piece being cut becomes very small. Consequently, the device envisaged in WO 99/44583 is only practical to make in diameters of greater than 1 mm. Implantation of such a large profile device requires a very sizable needle or trochar that would be very painful and cause significant local trauma to a patient. An additional problem with this device design is that it would be difficult to economically manufacture in large quantities. The reason is that each piece of sponge would need to be individually cut and stuffed into the tube. This would be very difficult to mechanize and perform rapidly.
Accordingly, it would be advantageous to provide an implantable device and method for modulating an immune response to specific antigens in mammals, similar in concept to the design described in WO 99/44583, whose filling preserves the porosity presented by a porous sponge, which is essential for rapid cellular infiltration, yet overcomes the mechanical frailties of a sponge.